pwython

API + RPC

What pests said. And you can make a preset and pass "model": "@preset/deep-seek"

Is your system set to dark mode? These days I find Claude often generates pages (like reports) with light & dark mode styles, and rightfully defaults to system preference.

Sometimes you want a decent model running in the background that doesn't take up all the VRAM.

> is there a name for this?

https://en.wikipedia.org/wiki/Open-core_model

I feel like this could be useful for my Karpathy-style LLM Wiki, but then again I think, in this day and age most of what I'm reading about is outdated in a few weeks. Not only are the tools/methods coming and going, but the ones that stick around drastically change. It would probably turn into a dump of antiquated markdown that could cause confusion.

As far as the need for private search, well, I've already searched for or visited those pages, so...

Goodnotes is already doing it. Their web app runs a large amount of the same Swift code used by their iOS app inside the browser through WebAssembly.

"This technology can be abused" and "this technology provides no public-safety value" are two entirely different claims.

Saying their ability to solve crimes is "limited to the humans using the tools" isn't much of an argument against the technology. That's true of fingerprints, DNA databases, radios, firearms databases, and virtually every other investigative tool.

A police officer can misuse a DMV database to stalk an ex. That means the officer should be prosecuted and access to the database tightly controlled. It doesn't demonstrate that DMV records have no legitimate investigative purpose. The same reasoning applies to LPRs.

By the way, that mom that was arrested? The kid was 5 years old, and was intentionally sent to a pond.

Honest question: If the feds were actually tracking you on your way to church, how does that impact you? Is that a real fear you have?

I'm 100% in favor of LPR cameras. Any technology can be abused. The handful of corrupt officers that were caught using the cameras for nefarious reasons shouldn't overshadow the incredible benefit these are.

> Willing to [bet] most of the Flock hate crowd doesn't...

...have a reason at all to worry about LPR cameras tracking their car. Just "something something government."

In OpenAI's Hugging Face report, they said that during training, agents "would first write notes into shared infrastructure, often as a form of external memory or to test some underlying system. When other agents came across these artifacts, it sometimes led them to infer that other agents were present."

They then give what they call a "hypothetical example but exemplary" of messages encoded in URL paths on a shared index page: "agent-07: answer(Q12)=42; need answer(Q19)=?".

So that's a GET request being used to pass information back and forth across multiple rounds. That's basically the DSEWiki pattern exactly. They say this likely came from the agents generalizing what they had learned from training with the official multi agent collaboration tool.

The report called it "misaligned behavior in an outlier scenario involving a rare and unexpected confluence of events."

They never mentioned a wiki, but this is most certainly it.

Is there any reason to even use 3.1 Pro now?

This is why I created: https://github.com/alisorcorp/warrant

It keeps documentation from going out of date by embedding re-runnable verification checks directly inside markdown files (that the agents use, not humans). I use it with a handoff workflow to force Claude/GPT to re-verify facts before handing off documentation to the next session.

Yea Qwen3.8 wasn't fun to use on my 64GB M4 Max either (better than these numbers though), so my new daily driver is Ornith-1.5-35B-A3B-MLX-4bit. I recommend giving that a whirl if you're on similar hardware, it's definitely better than Qwen3.6 35b-a3b which was my go-to before.

https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B

Sora 2 was way better than Veo in terms of quality. It just didn't have much of a use case for businesses outside of one-shot zany UGC-looking clips.

https://youtu.be/QaiecWzeHFM?si=UV6eF-m514nCOFe0&t=156

There's over 3 million models on HuggingFace. So many quants. I'd imagine less than 0.5% would actually be seeded.

Sure there is. You've been able to search/sort/filter/group for a while now.

I was already rolling around the idea of a 128GB M5 Max MBP. Now this!

A 4-bit MLX quant with 128k window should fit perfectly, in the 50-70 tok/s range.

I have actual obscure RGB corner floor lamps with BLE in my office. I had Claude find the BLE protocol and use my Office 365 calendar feed to change their color to red when I'm in a scheduled meeting (event), and back to warm white when the meeting is over, so co-workers know I'm busy without doing "are you busy" sign language through my window. The script is polled every 60 seconds. Tinkering with stuff like this is a lot of fun.

Very interesting, I missed your original post[0], will definitely look into it further. What's new since you've posted?

[0] https://news.ycombinator.com/item?id=48631726

What are the benefits? What would/could I do with the results?

Yep, terrible speller -- bad at pronunciation too (by the way I just autocorrected that word) -- and I often wonder how people, who are much more intelligent than me on paper, can't figure out random tasks that just seem obvious.

Yes. "Weak detail recall but strong fluid reasoning, conceptual/gist memory, and adaptive problem-solving" is how GPT sums up what I feel. And "high fluid intelligence" is a label it gave that I won't be using anytime soon... But I like the sound of it.

I hear ya, but one could set different operating thresholds: auto-approve low-risk posts, hold ambiguous posts for review, and automatically reject very high-confidence violations. So HITL for sure, but MUCH less H in the L.

I've had dreams of building something in the image sharing or social platform realm, but stopped short of planning because of obvious content moderation responsibilities. This seems to be a realistic, cost effective solution to that one piece of the puzzle.

H3 results are underwhelming compared to LTX 2.3 so far:

https://www.reddit.com/r/StableDiffusion/comments/1vciy35/lt...

Perhaps refined ComfyUI workflows will squeeze more quality out of it, but it's definitely not in the realm of Seedance, and Lightricks is training LTX 2.5/"LTX-Next".

Can't VS Code extensions solve your use case?

I was on a Royal Caribbean ship a couple weeks ago (Utopia, 18 decks, 5,668 passengers). It has 24 main elevators split between front and back, left and right. So a bank of 6 in front of you each time you needed a lift. You press your floor on a touchpad and it assigns you an elevator A through F.

I swear no matter if it was busy or not, it seemed like everyone was assigned to the same one or two elevators. The wait was always so long too, I used the stairs most of the time if just going <6 flights.

Someone besides RCCL needs to study that algorithm.

I use the Claude Code desktop app for work, and the CLI for personal, so I'm always logged into both accounts!

For house keeping tasks the robotics companies love showing off... it honestly doesn't matter if it takes a robot longer than a human to clean your house. As long as it gets done before you get home from work, no big deal.